Re: [Qemu-devel] [PATCH for-3.0 1/9] migration: simplify check to use qemu file buffer

[Peter Xu]

On Thu, Jul 05, 2018 at 10:01:09AM +0100, Dr. David Alan Gilbert wrote:
> * Peter Xu (address@hidden) wrote:
> > Firstly, renaming the old matching_page_sizes variable to
> > matching_target_page_size, which suites more to what it did (it only
> > checks against target page size rather than multiple page sizes).
> > Meanwhile, simplify the check logic a bit, and enhance the comments.
> > Should have no functional change.
> > 
> > Signed-off-by: Peter Xu <address@hidden>
> > ---
> >  migration/ram.c | 17 +++++++++++------
> >  1 file changed, 11 insertions(+), 6 deletions(-)
> > 
> > diff --git a/migration/ram.c b/migration/ram.c
> > index 23cea47090..fbeb23f750 100644
> > --- a/migration/ram.c
> > +++ b/migration/ram.c
> > @@ -3580,7 +3580,7 @@ static int ram_load_postcopy(QEMUFile *f)
> >  {
> >      int flags = 0, ret = 0;
> >      bool place_needed = false;
> > -    bool matching_page_sizes = false;
> > +    bool matching_target_page_size = false;
> >      MigrationIncomingState *mis = migration_incoming_get_current();
> >      /* Temporary page that is later 'placed' */
> >      void *postcopy_host_page = postcopy_get_tmp_page(mis);
> > @@ -3620,7 +3620,7 @@ static int ram_load_postcopy(QEMUFile *f)
> >                  ret = -EINVAL;
> >                  break;
> >              }
> > -            matching_page_sizes = block->page_size == TARGET_PAGE_SIZE;
> > +            matching_target_page_size = block->page_size == 
> > TARGET_PAGE_SIZE;
> 
> That's OK, although 'matches_target_page_size' would be better wording.
> ('matching_target_page_size' implies we know something about the sources
> target page size)

Sure thing.

> 
> >              /*
> >               * Postcopy requires that we place whole host pages atomically;
> >               * these may be huge pages for RAMBlocks that are backed by
> > @@ -3668,12 +3668,17 @@ static int ram_load_postcopy(QEMUFile *f)
> >  
> >          case RAM_SAVE_FLAG_PAGE:
> >              all_zero = false;
> > -            if (!place_needed || !matching_page_sizes) {
> > +            if (!matching_target_page_size) {
> > +                /* For huge pages, we always use temporary buffer */
> >                  qemu_get_buffer(f, page_buffer, TARGET_PAGE_SIZE);
> >              } else {
> > -                /* Avoids the qemu_file copy during postcopy, which is
> > -                 * going to do a copy later; can only do it when we
> > -                 * do this read in one go (matching page sizes)
> > +                /*
> > +                 * For small pages that matches target page size, we
> > +                 * avoid the qemu_file copy.  Instead we directly use
> > +                 * the buffer of QEMUFile to place the page.  Note: we
> > +                 * cannot do any QEMUFile operation before using that
> > +                 * buffer to make sure the buffer is valid when
> > +                 * placing the page.
> >                   */
> >                  qemu_get_buffer_in_place(f, (uint8_t **)&place_source,
> >                                           TARGET_PAGE_SIZE);
> 
> But, yes I think that's right; since this is in ram_load_postcopy
> the only time we have !place_needed is in the not-last subpage of a 
> hugepage (or host page > target_page), so that's already covered
> by the !matching_page_sizes check.
> 
> 
> Reviewed-by: Dr. David Alan Gilbert <address@hidden>

Thanks!

-- 
Peter Xu